It’s only a truth of life. A BJT wired in frequent emitter, even after compensating for the results of machine and temperature variations, nonetheless isn’t an ideal present supply.
Wow the engineering world together with your distinctive design: Design Concepts Submission Information
One of many flaws within the ointment is the Early impact of collector voltage on collector present. It could actually typically be estimated from datasheet parameters if output admittance (hoe) is specified (Ee ~ hoe / check present). A consultant worth is 1% per volt. Determine 1 exhibits its mischief in motion within the habits of a easy present mirror, the place:
I2 = I1(1 + Vcb/Va)
Va ~ 100v
Ierr = Vcb/Va ~ 1%/V
Determine 1 Present mirror with out emitter degeneration.
If the 2 transistors are matched, I2 ought to equal I1. However as a substitute, Q2’s collector present might enhance by 1% per Vcb volt. A double-digit Vcb might create a double-digit share error. That may make for a reasonably foggy “mirror”!
Thankfully, a easy trick for mitigating Early is well-known to expert practitioners of our artwork. (Please see the footnote). Emitter degeneration is predicated on an impact that’s 4000 instances stronger than the impact of Vcb on Ic.
That’s the impact of Vbe on collector present, and it may simply scale back Ee by two orders of magnitude. Determine 2 exhibits the way it works:
I2 ~ I1(1 + Vcb/Va) – (0.026R)(I2 – I1))
Ierr ~ (Vcb/Va)/(Vr/26mV + 1)
Determine 2 Present mirror with emitter degeneration
Equal resistors R added in collection with each emitters will develop voltages Vr = I1*R and I2*R that will likely be equal if the currents are equal. But when the currents differ (e.g., due to Early), then a Vbe differential will seem…duh…
That is helpful as a result of the Vbe differential will oppose the preliminary present differential, and the impact is giant, even when Vr is small. Determine 3 exhibits how dramatically this reduces Ierr.
Determine 3 A normalized Early impact (y-axis) versus emitter degeneration voltage Ve = Ia*R (x-axis). Observe that simply 50 mV reduces Early by 3:1. That’s certainly a “good distance”!
Footnote
One DI has an earlier dialog about present mirrors and the Early impact: “A two-way mirror—present mirror that’s.” Within the grand custom of editor Aalyia’s DI kitchen, frequent and knowledgeable commentator Ashutosh instructed how emitter degeneration may enhance efficiency:
asa70
Might 27, 2025
Relating to degen, i’ve discovered {that a} half volt at say 1mA FS helps match the out and in currents a lot better even at a tenth of the present, even for completely randomly chosen transistors. I suppose it’s as a result of the curves will likely be nearer at smaller currents, in order that even a 50 mV drop goes a good distance
Ashutosh actually nailed it! 50mV does go a protracted (3:1) approach!
Stephen Woodward’s relationship with EDN’s DI column goes again fairly a good distance. Over 100 submissions have been accepted since his first contribution again in 1974.
Associated Content material
- A two-way mirror—present mirror that’s
- A two-way Wilson present mirror
- Use a present mirror to regulate an influence provide
- A comparability between mirrors and Corridor-effect present sensings
The put up A present mirror reduces Early impact appeared first on EDN.
